The aim of this study of a blood-perfused isolated rabbit heart preparation was to differentiate the effects on coronary resistance of large changes in arterial O2 tension (arterial PO2 = 45-400 Torr) from the effects of variations in arterial O2 content or myocardial O2 delivery. Standard stored human blood was resuspended in Krebs-Henseleit buffer and was oxygenated to obtain normal PO2, high PO2, and low PO2. Hemoglobin concentrations were adjusted to obtain the same arterial O2 content (CaO2) for the three PO2s. In a first set of experiments, in which coronary blood flow (CBF) was free and adapted to a constant perfusion pressure, switching from control [138 +/- 17 (SE) Torr] to high PO2 blood (380 +/- 27 Torr) induced a significant decrease in CBF and myocardial O2 consumption (MVO2). Switching from control (125 +/- 3 Torr) to low PO2 blood (49 +/- 5 Torr) induced a significant increase in CBF and MVO2. In a second set of experiments, the switch from control (159 +/- 5 Torr) to high PO2 (389 +/- 32 Torr) was performed in a preparation in which CBF and consequently O2 delivery were constant. Under these conditions, the increase in perfusion pressure demonstrated that PO2 affected coronary resistance, even though the O2 delivery was constant. No significant change in myocardial performance was observed in any of these experimental procedures. These results show that arterial PO2 may affect coronary blood flow regulation independently of any mediation by the autonomic nervous system and of any associated changes in O2 content or O2 delivery.